Article pubs.acs.org/biochemistry
Molecular Level Interaction of Inositol Hexaphosphate with the C2B Domain of Human Synaptotagmin I Meng-Je Joung, Sepuru K. Mohan, and Chin Yu* Department of Chemistry, National Tsing Hua University, Hsinchu, Taiwan ABSTRACT: Synaptotagmin I is a synaptic vesicle membrane protein that serves as a multifunctional regulator during the exocytosis of neurotransmitter release. It contains C2A and C2B domains. The binding of Ca2+ to the C2A domain activates the exocytosis of secretory vesicles, while the binding of inositol polyphosphates (IP4−IP6) to the C2B domain inhibits this process. To understand the IP6induced inhibition of exocytosis of secretory vesicles, we determined the threedimensional structure of the C2B−IP6 complex by nuclear magnetic resonance (NMR). In this study, we have determined the binding constant by isothermal titration calorimetry. The circular dichroism measurements demonstrated that IP6 can stabilize the C2B molecule. We identified the binding site using 1H−15N heteronuclear single-quantum coherence spectroscopy titration data and determined the structure of the C2B−IP6 complex using multidimensional NMR studies. This information will aid in the design of better pharmacological treatments for neurological disorders.
S
ynaptotagmins (Syts) make up a family of vesicle membrane proteins that includes more than 12 isoforms with diverse functions and tissue-specific expression patterns.1,2 Syts all contain a short N-terminal intravesicular transmembrane region and tandem cytoplasmic repeats that are homologous to the C2 regulatory region of C2A and C2B.3 Synaptotagmin I (Syt I), the best-characterized isoform, is expressed abundantly in neurons and is essential for fast Ca2+triggered neurotransmitter release.4 The C2A domain of Syt I is considered to be a calcium sensor, because the binding between phospholipids and the C2A domain is Ca2+-dependent.5 The C2A domain also binds to syntaxin, a plasma membrane protein, in the presence of Ca2+.6,7 This interaction is necessary for exocytosis of neurotransmitter release. Several proteins, such as clathrin assembly protein-2 (AP-2),8 soluble NSF attachment protein (β-SNAP),9 SNAP-25, and Ntype calcium channels,10−13 bind to different sites in the C2B domain. AP-2 is a multimeric protein complex that participates in cargo protein internalization during the clathrin-mediated endocytosis of synaptic vesicles on the plasma membrane. The interaction between the C2B domain and AP-2 was reported to play an important role in synaptic vesicle endocytosis. An in vivo study in Caenorhabditis elegans suggests that the C2B domain may function as a high-affinity docking site for AP-2 and serve as a bridge between endo- and exocytosis in the synaptic vesicle cycle.14 It was also reported that the C2B domain binds inositol polyphosphates [IHPS, including inositol 1,3,4,5-tetrakisphosphate (IP4), inositol 1,3,4,5,6-pentakisphosphate (IP5), and inositol hexakisphosphate (IP6)] and phosphoinositide polyphosphates (PtdInsPn).15−18 This binding is Ca2+ concentration-dependent. The binding module was determined using a liposome model system.19 © 2012 American Chemical Society
IP6 is generally considered to be an antinutrient because of its ability to chelate divalent minerals and reduce the extent of their absorption.20 IP6 was shown to inhibit free radical formation and weaken the lipid peroxidation catalyzed by iron and ascorbic acid in human erythrocytes.20,21 A protective effect for IP6 was suggested because of its antioxidant effects and its ability to alter the cell signaling pathways that detoxify ROS (reactive oxygen species). A cell line assay using squid giant synapses indicated that the serial microinjection of IP6 into the presynaptic terminal inhibited synaptic transmission; this inhibition could be released by the co-injection of antibodies that recognize the C2B domain.22 In addition, IP6 also plays an essential role in regulating the inhibition of neurotransmitter release, and this effect might partially be due to the inhibition of the interaction between Syt I and AP-2, which blocks endocytosis.23 The relationship between protein structure and function is important for the development of pharmacological agents for the treatment of human diseases. The C2B domain of Syt I is potentially involved in synaptic vesicle exocytosis. The aim of this study is to understand the interactions between C2B and IP6 at the molecular level. Here, we have used a variety of biophysical methods, including isothermal titration calorimetry (ITC), circular dichroism (CD) spectrometry, and multidimensional NMR, to characterize the interactions between C2B and IP6. Received: January 2, 2012 Revised: March 8, 2012 Published: April 4, 2012 3675
dx.doi.org/10.1021/bi300005w | Biochemistry 2012, 51, 3675−3683
Biochemistry
■
Article
MATERIALS AND METHODS Reagents. GST-Sepharose was purchased from Amersham Pharmacia Biotech. 15NH4Cl, 13C-labeled glucose, and D2O were purchased from Cambridge Isotope Laboratories. The components for the Luria broth media were obtained from AMRESCO. The Centricon and Amicon membranes were purchased from Millipore. All of the other chemicals used in the study were of high-quality analytical grade. Protein Expression and Purification. The cDNA encoding the human C2B domain of Syt I (residues 270− 421) was subcloned into the pGEX-4T1 expression vector. Escherichia coli cells expressing GST-tagged C2B were induced with isopropyl β-D-thiogalactopyranoside (IPTG) when the absorbance (at 600 nm) of the culture reached 0.6−0.9. The cells were harvested by centrifugation at 6000 rpm after being induced for 16 h at 25 °C. The unlabeled protein was expressed in Luria broth (LB) medium. The harvested cells were resuspended in resuspension buffer [50 mM Tris-HCl (pH 7.4) and 300 mM NaCl], and the cell walls were ruptured with a French press and via sonication. The cell lysate was centrifuged at 16000 rpm for 30 min. The supernatant was then incubated with glutathione Sepharose, and the resin was then extensively washed with resuspension buffer and 50 mM Tris (pH 7.4) containing 1 M NaCl to remove impurities. The bound GST−C2B protein was eluted with 10 mM glutathione in cleavage buffer [50 mM Tris, 0.1 M NaCl, and 2.5 mM CaCl2 (pH 8.0)]. Thrombin cleavage was performed (2 NIH units/mL) at 25 °C for 20−24 h. Then, the whole solution was concentrated by Amicon and further purified by gel filtration on a Superdex 75 (Pharmacia) column using an FPLC system with 20 mM MES (pH 6.0) containing 150 mM NaCl and 2 mM DTT as the eluent. The purity of the protein was checked by sodium dodecyl sulfate−polyacrylamide gel electrophoresis, and its molecular weight was confirmed by electrospray mass analysis. The complete removal of nucleotide contaminants was verified by recording the UV spectrum of the protein. Preparation of 15N- and 13C-Labeled C2B. Uniform labeling of the C2B domain using isotopes 15N and 13C was achieved by growth in M9 minimal medium. This medium contained either 15NH4Cl for single (15N) labeling or 15NH4Cl and [13C]glucose for double (15N and 13C) labeling of the C2B domain. The maximal expression yield was achieved using a modified M9 medium that included additional vitamins. The host expression strain, E. coli BL21(DE3), is a vitamin B1deficient host, and therefore, thiamin (vitamin B1) was added to achieve yields of up to 10−15 mg of protein per liter from cells grown in the isotope-enriched medium. Isothermal Titration Calorimetry (ITC). Protein−ligand binding was characterized by measuring the heat changes during the titration of an interacting ligand into a protein solution using a Microcal VP titration calorimeter. C2B and IP6 solutions were centrifuged and degassed under vacuum before being used. Titrations were performed by injecting 8 μL aliquots of IP6 (30 times, 1 mM) into a 0.1 mM C2B solution. The titrations were performed at 25 °C, with the protein and ligands dissolved in 20 mM MES (pH 6.0) containing 150 mM NaCl, 2 mM DTT, and 2 mM CaCl2. The results of the titration curves were corrected using a buffer−protein control. Initially, the buffer alone was titrated, using IP6 as the ligand. Later, the same procedure was repeated with the protein dissolved in the buffer. The heat change of the first experiment (buffer titrated with IP6) was subtracted from that of the
second (protein dissolved in buffer titrated with IP6), as shown in Figure 1. Finally, the Origin software supplied by Microcal was used to analyze these data. CD Measurements. Circular dichroism measurements (CD) were taken using an AVIV CD spectrophotometer to determine the thermal stability of the protein. The thermal denaturation curves of the native free C2B domain and the C2B−IP6 complex were determined over a temperature range of 23−90 °C, at intervals of 3 °C for each data set, to determine the Tm of the protein in the presence and absence of the ligand. The protein sample was dissolved in 20 mM MES buffer (pH 6.0) containing 150 mM NaCl, 2 mM DTT, and 2 mM CaCl2. The concentration of the protein sample was 10 μM, and the C2B−IP6 complex was present at a 1:1 C2B domain:IP6 ratio. The CD spectrum of the free C2B domain shows a positive ellipticity band at 228 nm, which is characteristic of a β-barrel protein, while the denatured protein does not produce a 228 nm CD band. The ΔG of unfolding was calculated from the slope of the folded and unfolded state baselines. NMR Experiments. All of the two-dimensional (2D) and three-dimensional (3D) NMR resonance experiments were performed using a Varian 700 MHz NMR spectrometer equipped with a cold probe at 25 °C. The protein sample was present at 1.0 mM in a solution containing 20 mM MES (pH 6.0), 150 mM NaCl, 2 mM DTT, and 2 mM CaCl2 in the presence of 10% D2O. In a 13C-filtered NOESY experiment, the spectra of the protein complex were recorded in 100% D2O. The 15N-labeled protein was titrated with IP6 at a 1:1 molar ratio in the binary complex. A plot of the weighted average of the 15N and 1H chemical shift perturbations of the protein residues was calculated using the equation Δδ = [(δ1H)2 + 0.2(δ15N)2]1/2. All of the spectra were processed using VNMR and analyzed using Sparky.24 3D NMR Experiments. The C2B resonances in the C2B− IP6 complex were assigned using various multidimensional NMR experiments. The backbone 1H, 13C, and 15N resonances in the complex were assigned via 3D HNCA and HNCOCA experiments.25 The side chain resonances were assigned using 3D 15N-edited TOCSY-HSQC, HCCH-COSY, and HCCHTOCSY data sets supplemented with data from other experiments, including CBCA(CO)NH26 and HBHA(CO)NH.27 In addition, the Cα and Cβ resonances were identified by CBCA(CO)NH experiments, and the carbon side chains of the protein binary complex were resolved by the CC(CO)NH experiment. HNCO spectra were used to assign the carbonyl carbons.28 We also used the 3D 15N-edited NOESY spectrum to identify the NH2 groups of the amino acids Gln and Asn. The aromatic resonances of C2B were assigned using simultaneous 13C- and 15N-edited NOESY-HSQC spectra.29 Intermolecular distance restraints were derived from the 3D 13 C, 15N(F1)-filtered, 13C(F2)-edited, and 12C(F3)-filtered NOESY-HSQC spectrum30 of a 1:1 15N-, 13C-, and 1H-labeled C2B/IP6 (unlabeled) mixture. Structure Calculation. The structure of the C2B complex was calculated iteratively using ARIA/CNS (version 2.2) with the PARALLHDG 5.3 force field in the PARALLHDG mode.31 We used a variety of triple-resonance NMR experiments to determine the solution structure of C2B in the C2B−IP6 complex. Interproton distance restraints were derived from the NOE cross-peak intensities in the NOESY spectrum. In addition, hydrogen bond information was derived from a hydrogen−deuterium exchange experiment, and dihedral angle restraints were generated using TALOS32 with the HN, CA, 3676
dx.doi.org/10.1021/bi300005w | Biochemistry 2012, 51, 3675−3683
Biochemistry
Article
CB, CO, HA, and HB atoms. A total of 200 structures were calculated and further refined with CNS in an explicit solvent layer of water from which the best 16 structures with the lowest energies were selected. Here, we used PROCHECK to analyze the structure of the protein complex and extract its structural parameters.33 MOLMOL and PYMOL were then used to generate structural representations. Molecular Docking. The C2B−IP6 complex was assessed using HADDOCK (high-ambiguity-driven docking)34−37 in combination with CNS.38 The IP6 coordinates were taken from the Protein Data Bank (PDB) (entry 2K8R). The proton resonances of free IP6 were assigned using one-dimensional and 2D TOCSY and 2D NOESY, and the proton resonances of IP6 bound to C2B were assigned using an analysis of isotopefiltered TOCSY and isotope-filtered NOESY spectra.39 The topology and the parameter files were generated using the HICUP server.40 The docking procedure was driven by the intermolecular data derived from the 3D 13C, 15N(F1)-filtered, 13 C(F2)-edited, and 12C(F3)-filtered NOESY-HSQC data, which were used as restraints to dock IP6 with the ARIA/ CNS-derived structure of the C2B domain. The intermolecular NOE data were used as unambiguous restraints, and the residues that showed chemical shift perturbations were used to define the ambiguous interaction restraints (AIRs) for each residue at the interface region as either active or passive. NACCESS41 was used to identify the solvent-exposed residues in the C2B domain. Active residues had solvent surface accessible areas of >50%, while the passive residues had solvent surface accessible areas of 0.1 ppm; however, residues K313, H315, L323, K325, K326, K327, G368, D371, K375, and V376 had the greatest chemical shift perturbations (Figure 2B). Residues displaying large chemical shift perturbations are
■
RESULTS AND DISCUSSION Syt I functions as a major Ca2+ sensor in neurotransmitter release through the binding of Ca2+ to its two C2 domains.43,44 The two C2 domains have similar overall structures but different functions; binding of Ca2+ to the C2A domain serves a regulatory role in release, but binding of Ca2+ to the C2B domain is essential for release. However, it has been reported that IP6 strongly bound to the C2B domain and inhibited the fusion step of Ca2+-regulated exocytosis.45,46 Therefore, IP6 was assumed to be a potential regulator of neurotransmitter release. Thus, the elucidation of the structural interactions between the IP6 and C2B domains at the molecular level would provide the mechanistic information needed to understand how IP6 acts as an inhibitor of neurotransmitter release. In this study, we focused on determining the solution structure of the C2B−IP6 complex. Isothermal Titration Calorimetry (ITC). The ITC technique measures the heat change associated with binding by simply titrating the ligand into a solution containing the macromolecule.47 The heat changes are integrated and fit to obtain the full set of thermodynamic parameters of an interaction. Initially, injections of IP6 into the solution containing C2B domains produced strong heat changes. With progressive injections, the magnitude of the heat signal 3677
dx.doi.org/10.1021/bi300005w | Biochemistry 2012, 51, 3675−3683
Biochemistry
Article
Figure 3. Analysis of the thermal stability of the C2B domain in the presence (■) and absence (◆) of IP6. Changes in the far UV-circular dichroism spectra were recorded by measurement of the ellipticity as a function of wavelength at 228 nm. The CD spectra were recorded in 20 mM MES buffer (pH 6.0) containing 150 mM NaCl, 2 mM DTT, and 2 mM CaCl2 using a 1 mm cylindrical cuvette. The concentration of the protein sample was 10 μM, and the C2B−IP6 complex included a 1:1 C2B domain:IP6 ratio.
Table 1. Structural Statistics of the C2B Domain in the C2B−IP6 Complex from the ARIA/CNS Structure Calculation of the 20 Best Conformers no. of distance constraints long-range medium-range short-range intraresidue (i = j) no. of dihedral angle constraints (φ/ψ) no. of hydrogen bond constraints no. of intermolecular NOEs average rmsd from the mean structure (Å) residues 3−131 backbone atoms heavy atoms regular secondary structure elements backbone atoms heavy atoms deviations from idealized geometry bond lengths (Å) bond angles (deg) impropers (deg) Procheck G factors dihedrals covalent overall Ramachandran statistics (% of all residues) most favored additionally allowed generously allowed disallowed
Figure 2. Analysis of the free C2B domain and the C2B−IP6 complex using 2D NMR at a 1:1 binding ratio. (A) The overlaid 2D 1H−15N HSQC spectra highlight the spectral changes in the uniformly 15Nlabeled C2B domain (green) and the C2B domain upon binding to IP6 (red). (B) Weight average of the chemical shift (1H and 15N) perturbations {Δδ = [(δ1HN)2 + 0.2(δ15N)2]1/2} of the amino acid residues in the C2B domain upon complex formation with IP6. The inset depicts the significant chemically shifted residues mapped over the ribbon diagram of the C2B structure.
typically thought to indicate the putative binding region of the ligand. However, these pretreated residues were located in two distinct regions of the C2B domain (Figure 2B, inset). In Figure 2B, residues K313, H315, L323, and K325−K327 are colored yellow and G368, D371, K375, and V376 are colored red. IP6 was bound around these residues, except K375 and V376. These two perturbed residues (K375 and V376) might be chemically shifted because of secondary effects. It has been reported that point mutants in the polybasic Lys residues (K324−K327) within the C2B domain decrease the binding affinity of the C2B domain for IP6.49 In addition, the ITC data suggest a one-binding site mode for the C2B−IP6 complex. To determine the exact binding region of IP6 on C2B, we determined the solution structure of the C2B−IP6 complex using intermolecular NOE data and chemical shift perturbation data. The C2B Domain Is Stabilized by the Binding of IP6. The conformational changes that occur during the process of thermal denaturation in a protein can be determined by CD
918 411 462 396 110/110 60 4
0.91 ± 0.12 1.26 ± 0.11 0.56 ± 0.09 1.06 ± 0.11 0.005 ± 0.001 0.461 ± 0.056 0.585 ± 0.081 −0.21 0.71 0.03 85.0 12.6 1.6 0.8
spectroscopy. The thermal denaturation of both the free C2B domain and the IP6-bound form of the C2B domain was analyzed at temperatures ranging from 25 to 85 °C. The comparison and analysis of the CD data revealed that the melting temperature (Tm) of the C2B domain increased by 3 °C when it was bound to IP6 (Figure 3), as indicated by the 3678
dx.doi.org/10.1021/bi300005w | Biochemistry 2012, 51, 3675−3683
Biochemistry
Article
Figure 4. Solution structure of C2B in the C2B−IP6 complex. (A) Overlay of 16 structures showing the backbone representation of the C2B domain in the C2B−IP6 complex. The eight β-strands are colored cyan; the two α-helices are colored red, and the loop region is colored gray. (B) Ribbon representation of the C2B domain in the C2B−IP6 complex (average structure). (C) Stereopair showing C2B in the C2B−IP6 complex.
were subsequently generated from the 15N-edited NOESYHSQC and 13C-edited NOESY-HSQC data and were used for the structure determination calculations. Hydrogen bond restraints derived from the hydrogen−deuterium exchange experiment were also used for structure determination. The structural restraints and statistics for the ensemble of the NMR structures calculated with ARIA/CNS are listed in Table 1. A total of 2187 NOEs were used for the calculation of the C2B− IP6 binary complex structure. The calculated structure of the C2B domain in the C2B−IP6 binary complex agrees well with the NMR structure of the free C2B domain (PDB entry 1K5W) in the structured region (two α-helices and eight βsheets), with a backbone rmsd of 0.85 Å (Figure 4). According to a Ramachandran analysis, 85% of the residues were in the most favored region and